Giacomo Fiorin

Reading more about the format from the SciPy page and other open source libraries like this one: https://github.com/tbeu/matio It seems that since version 7.3 (released in 2006) the format is...

That's along the lines of what I meant in the previous comments: the state file has all that information encoded. The scripting interface could easily provide access to the grid...

@jhenin I pinged @fabsugar in person and he agrees with switching format for dimensions 3 and above. One issue I didn't think about immediately when you wrote your last comment...

I checked, in the `colvargrid::read_restart()` function there is this condition: https://github.com/Colvars/colvars/blob/7f850e071b2d34d5a54f872592af064a0841dbe8/src/colvargrid.h#L959-L966 that mimics a previous check first introduced in 2009 (before the start of the Git history), when the state...

BTW @HanatoK I rememberd that there is also [this case](https://github.com/Colvars/colvars/pull/447) of inputting data through a grid. I hope it's fine to keep the format as multi-column, otherwise a DX parser...

@fhh2626 Sorry, we had agreed to stop tracking the NAMD core files but I forgot to disable the script that keeps them up to date in the Colvars repository nightly....

Can you please try with clean worktrees for both Colvars and NAMD and using the `devel-tools/compile-namd.sh` provided in this repo?

Same here, the error is from an outdated build recipe from Lepton that misses a new file that they added upstream. But the Colvars patch script has been updating this...

How did you patch NAMD? If you use something different from the procedure listed here, please include it in detail.

@fhh2626 Please be more complete. We have already tried several times and the procedure used in the CI: https://github.com/Colvars/colvars/blob/da4ccd8af334d3b3f831915e7042ae6216fe268e/.github/workflows/backend-template.yml#L144-L147 works offline as well. Can you post here a link to...